{"ID":2921115,"CreatedAt":"2026-06-02T02:42:49.606572591Z","UpdatedAt":"2026-06-04T06:21:04.369492701Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.01816","arxiv_id":"2606.01816","title":"Site4Drug: Predicting Drug-Binding Target Sites with an AI Agent","abstract":"Selecting where to intervene on a protein (i.e., choosing a targetable site) is often a more ambiguous and failure-prone bottleneck than selecting what binds, especially for membrane proteins where accessibility, topology, and post-translational modifications (PTMs) constrain actionable regions. We present Site4Drug, a modality-aware site-finding agent that outputs a ranked list of targetable regions with explicit constraints, evidence summaries, risk flags, and a traceable decision log. Rather than requiring users to specify the drug modality upfront, Site4Drug can recommend a binding modality (e.g., antibody/peptide-like vs small-molecule) from the same evidence used for site discovery, including topology, hydropathy, PTM propensity, disulfides, domain context, and sequence. Importantly, this evidence is applied consistently across modalities, including small-molecule pocket discovery, to avoid selecting chemically plausible but biologically occluded sites.","short_abstract":"Selecting where to intervene on a protein (i.e., choosing a targetable site) is often a more ambiguous and failure-prone bottleneck than selecting what binds, especially for membrane proteins where accessibility, topology, and post-translational modifications (PTMs) constrain actionable regions. We present Site4Drug, a...","url_abs":"https://arxiv.org/abs/2606.01816","url_pdf":"https://arxiv.org/pdf/2606.01816v1","authors":"[\"Taehan Kim\",\"Sarrah Rose Mikhail Leung\",\"Bharat Mekala\",\"Jeongbin Park\"]","published":"2026-06-01T07:32:02Z","proceeding":"q-bio.BM","tasks":"[\"q-bio.BM\",\"cs.LG\"]","methods":"[]","has_code":false}